Method and an apparatus for correcting distortion of binary received signals

Abstract

In a method of correcting a signal having a sequence of binary bits modulated on a carrier and followed by demodulation of the signal, and wherein a distortion has been impressed on the signal during passage through a transmission channel, the steps include setting initial threshold values in dependence of initial reception conditions and wherein the threshold values normally vary as a function of time with a bit duration, comparing the demodulated signal several times within the bit duration with the set threshold values, and correcting the initially set threshold values in dependence of later reception conditions following the initial reception conditions.

Having thus described the invention, what I claim as new and desire to be secured by Letters Patent is as follows: 
     
       1. An apparatus for correcting a signal for distortion by means of threshold value comparisons, said signal having a sequence of binary "1" and "0" bits modulated on a carrier, and wherein the distortion has been impressed on said signal during passage through a transmission channel, comprising in combination   signal demodulation means for demodulating the signal,   sampling means for sampling the demodulated "1" and "0" bits each an equal number of times during a bit duration, including hold means for holding the sampled signal substantially at the sampled value, and wherein the sampled "0" bits form initial threshold values, and the sampled "1" bits form demodulated signal values,   analog-to-digital converter means postcoupled to said sampling means for quantizing the sampled values,   a first memory for storing the quantized initial threshold values,   a second memory for storing the quantized demodulated values,   a threshold calculator for calculating from the stored and quantized demodulated signal values and from the stored and quantized initial threshold values revised threshold values,   a first comparator for comparing the quantized demodulated signal values with corresponding of said revised threshold values, and for deciding whether the sampled signal value is a "1" value or a "0" value,   quality factor determining means for determining the quality factor of a detected bit including storage means for storing a reference quality factor,   a second comparator connected to said storage means for arriving at a decision whether said quality factor of a detected bit exceeds or falls below said reference quality factor in dependence of the output of said first comparator, and of said storage means,   read only memory means having address locations with variable contents, and   means for changing the contents of said address locations in dependent of the decision of said second comparator, and for feeding a series of parameters to said treshold calculator in dependence of the changed contents of said address locations.     
     
     
       2. An apparatus as claimed in claim 1, wherein said quality factor determining means further comprise a demultiplexer for separating the output of said comparator into separate "1" and "0" sampled values,   a second counter for counting the difference between said "1" and "0" sampled values within each "1" bit, and within each "0" bit, respectively, and for feeding the count obtained to said second comparator.   
     
     
       3. An apparatus as claimed in claim 1, wherein said means for changing the contents of said address locations includes an address counter for accepting the output of said second comparator, and for counting the difference of the number of sampled values within each bit exceeding or falling below said reference quality factor, and wherein the contents of said address locations are changed in further dependence on the output of said address counter. 
     
     
       4. In a method of correcting a signal having a sequence of binary bits modulated on a carrier and followed by demodulation of said signal, so as to obtain a "1" channel and a "0" channel, and wherein a distortion has been impressed on said signal during passage through a transmission channel, the steps comprising   setting initial threshold values in dependence of initial reception conditions, said threshold values normally varying as a function of time within a bit duration,   selecting said threshold values within said bit duration so as to be a sequence of a predetermined number of discrete integer sampled threshold values S i ,j,m,   sampling the demodulated signal so as to obtain a prearranged number of discrete sampled values of the demodulated signal A i ,j,m, equal to said predetermined number, said sampled threshold values being determined as a function of the relation: ##EQU5##  wherein S i ,j,m-1 is a chosen discrete sampled threshold value of a bit preceding a corresponding chosen discrete sampled threshold value of a selected bit, K is a weighting constant stored in a permanent value memory, and A is an average value obtained from the values A i ,j,m and A i ,j,m-1 being a chosen discrete sampled value of the bit of the demodulated signal preceding a corresponding chosen discrete sampled value of a selected bit of the demodulated signal,   comparing the discrete sampled values A i ,j,m of the demodulated signal several times within said bit duration with the selected threshold values S i ,j,m, and   assigning to a received bit a logical "1" value when a majority of the discrete sampled values of the demodulated signal A i ,j,m within said received bit have a value exceeding the corresponding discrete sampled threshold values S i ,j,m in said "1" channel, and   assigning to a received bit a logical "0" value when a majority of said discrete sampled threshold values S i ,j,m within said received bit have a value exceeding the corresponding discrete sampled values of the demodulated signal A i ,j,m in said "0" channel, and thereafter   correcting the initially selected threshold values in dependence of later reception conditions following said initial reception conditions.   
     
     
       5. In a method of correcting a signal having a sequence of binary bits modulated on a carrier and followed by demodulation of said signal, and wherein a distortion has been impressed on said signal during passage through a transmission channel, the steps comprising   setting initial threshold values in dependence of initial reception conditions, said threshold values normally varying as a function of time within a bit duration,   comparing the demodulated signal several times within said bit duration with the initially set of threshold values,   selecting a quality factor (G) for each bit of said bit sequence,   comparing the absolute value of said quality factor with a positive reference quality factor (G Q ), and   correcting the initially set of threshold values in dependence of later reception conditions following said initial reception conditions.   
     
     
       6. The method as claimed in claim 5, further comprising the steps of increasing the address of a read only memory by one upon said quality factor (G) exceeding the absolute value of said reference quality factor (G Q ), of decreasing said address by one upon said quality factor (G) falling below said absolute value of said reference quality factor (G Q ), and storing in said read only memory falling or equal values of a weighting constant (K) upon said address being increased. 
     
     
       7. In a method of correcting a signal having a sequence of binary bits modulated on a carrier and followed by demodulation of said signal, so as to obtain a "1" channel and a "0" channel, and wherein a distortion has been impressed on said signal during passage through a transmission channel, the steps comprising   setting initial threshold values in dependence of initial reception conditions, said threshold values normally varying as a function of time within a bit duration,   selecting said threshold values within said bit duration so as to be a sequence of a predetermined number of discrete integer samples threshold values S i ,j,m,   sampling the demodulated signal so as to obtain a prearranged number of discrete samples values of the demodulated signal A i ,j,m, equal to said predetermined number, said sampled threshold values being determined as a function of a chosen discrete sampled threshold value S i ,j,m-1 of a bit preceding a corresponding chosen discrete sampled threshold value of a selected bit, of a weighting constant K stored in a read only memory, and of an average value A i ,j,m obtained from the values A i ,j,m and A i ,j,m-1, being a chosen discrete sampled value of the bit of the demodulated signal preceding a corresponding chosen discrete sampled value of a selected bit of the demodulated signal,   comparing the discrete sampled values A i ,j,m of the demodulated signal several times within said bit duration with the corresponding selected threshold values S i ,j,m,   assigning to the comparison a positive decision result when the discrete sampled values A i ,j,m of the demodulated signal within said received bit is equal to or exceeds the corresponding discrete sampled threshold values S i ,j,m,   assigning to a received bit a logical "1" value when the number (N 1 ) of positive decisions of the "1" channel is greater than the number (N 0 ) of positive decisions of the "0" channel,   assigning to a received bit a logical "0" value when the number (N 0 ) of positive decisions of the "0" channel is greater than the number (N 1 ) of positive decisions of the "1" channel, and   correcting the initially selected threshold values in dependence of later reception conditions following said initial reception conditions.   
     
     
       8. In a method of correcting a signal having a sequence of binary bits modulated on a carrier and followed by demodulation of said signal, so as to obtain a "1" channel and a "0" channel, and wherein a distortion has been impressed on said signal during passage through a transmission channel, the steps comprising   setting initial threshold values in dependence of initial reception conditions, said threshold values normally varying as a function of time within a bit duration,   selecting said threshold values within said bit duration so as to be a sequence of a predetermined number of discrete integer sampled threshold values S i ,j,m,   sampling the demodulated signal so as to obtain a prearranged number of discrete sampled values of the demodulated signal A i ,j,m, equal to said predetermined number, said sampled threshold values being determined as a function of the relation: ##EQU6##  wherein S i ,j,m-1 is a chosen discrete sampled threshold value of a bit preceding a corresponding chosen discrete sampled threshold value of a selected bit, K is a weighting constant stored in a read only memory, and A i ,j,m is an average value obtained from the values A i ,j,m and A i ,j,m-1, A i ,j,m-1 being a chosen discrete sampled value of the bit of the demodulated signal preceding a corresponding chosen discrete sampled value of a selected bit of the demodulated signal,   comparing the discrete sampled values A i ,j,m of the demodulated signal several times within said bit duration with the corresponding selected threshold values S i ,j,m,   assigning to the comparison a positive decision result when the discrete sampled values A i ,j,m of the demodulated signal within said received bit is equal to or exceeds the corresponding discrete sampled threshold values S i ,j,m,   assigning to a received bit a logical "1" value when the number (N 1 ) of positive decisions of the "1" channel is greater than the number (N 0 ) of positive decisions of the "0" channel,   assigning to a received bit a logical "0" value when the number (N 0 ) of positive decisions of the "0" channel is greater than the number (N 1 ) of positive decisions of the "1" channel, and   correcting the initially selected threshold values in dependence of later reception conditions following said initial reception conditions.